'Birth to Ten' - pilot studies to test the feasibility of a birth cohort study investigating the effects of urbanisation in South Africa

The report includes the findings of a pilot feasibility study in the Johannesburg/Soweto metropolitan area preliminary to a major longitudinal birth cohort study Birth to Ten (BT10). The objective was to develop appropriate sampling methods and to obtain information on the monthly birth rate the timing and frequency of antenatal visits and visits to wellbaby clinics and the availability of delivery data and the accuracy and reliability of routinely collected growth data. A questionnaire on birth data was developed and the logistics pretested in delivery centers; cord blood was also collected. An introductory section provides background on the urban dynamics of South Africa and identifies the issued in conducting a BT10 study. The following discussion focuses on the methods the results and discussion. Monthly births were obtained from local authority area records of registered births in hospitals and delivery centers and home deliveries of registered midwives. Not included are the 1987/8 estimated .54% of births attended to by nonregistered personnel in Johannesburg; no such data is available in Soweto. A random sample of registered births (not including private hospitals) in June 1987 was selected; mothers prenatal care from hospital records was obtained on the frequency timing and duration of care. Immunization data on all births were obtained from a pretest in Baragwanath Hospital and a satellite clinic; however standaridized growth data was unavailable throughout the entire area. The results for example on the monthly birth rate were 2580 births/month. The pilot sample included 619 births of which 470 were provincial delivery center births and 149 were private centers. Of the 470 mothers 80% received prenatal care and the average number of visits was 5. 89% were seen at least 2 times. Of the 619 births 394 were traced through records at wellbaby clinics. 88.8% received diphtheria pertussis tetanus and polio vaccinations. 83% received a measles vaccination. Delivery data had .9-23.6% of the data missing out of 21 items. 16% of hospital delivery data was missing perhaps because caesarean deliveries were routed differently. It was estimated from the pilot that a recommended sample size for a 6 week period would be 3350 single births of which 2010 could be followed up in 9 months. This figure satisfies statistical requirements and generalizability to the urban area. Measurement equipment and training would be necessary. The repeated discussions with health personnel increases the interest in the study and hence will increase effectiveness.